Researchers have found a new way to amplify trustworthy news content on social media

2022-05-07 0 By

Social media sites continue to amplify misinformation and conspiracy theories.To solve this problem, an interdisciplinary team of computer scientists, physicists, and social scientists led by the University of South Florida (USF) found a solution to ensure that social media users have access to more reliable news sources.In their study, published in the journal Nature Human Behavior, the researchers focused on recommendation algorithms that social media platforms use to prioritize content shown to users.Instead of measuring engagement by number of users and page views, the researchers looked at what was amplified in news feeds, focusing on the reliability scores of news sources and the diversity of their audiences.Low-quality content is appealing because it matches what we already know and like, whether it’s accurate or not.As a result, misinformation is often spread among like-minded viewers.The algorithm eventually picked the wrong signal and went on to generalize it further.To break this cycle, people should look for engaging content, but with a diverse audience, not like-minded people.Working with researchers from Indiana University and Dartmouth College, the team created a new algorithm using Web traffic data and self-reported party affiliations from 6,890 people, who reflect the diversity of the US in terms of gender, race and political affiliation.The data comes from YouGov, an online polling company.They also reviewed the reliability scores of 3,765 news sources based on the NewGuard Reliability Index, which rates news sources based on several journalistic criteria, such as editorial responsibility, accountability, and financial transparency.They found that combining the diversity of news audiences can improve the reliability of recommendation sources while still providing relevant recommendations to users.Because the algorithm is not entirely based on engagement or popularity, it can still promote reliable sources regardless of party affiliation.This is particularly welcome news for social media platforms, especially since they have been reluctant to make changes to their algorithms for fear of critical bias.The researchers say it is easy for platforms to incorporate audience diversity into their own recommendation algorithms because diversity measures can be derived from engagement data and are already recorded every time a user clicks a “like” or shares something on a news feed.Ciampaglia and his colleagues suggest that social media platforms adopt this new strategy to help prevent the spread of misinformation.